630 F. F. GROUT GRAPH TO STUDY OF IGNEOUS ROCK SERIES 



is called to it simply to stimulate careful field observations as to whether 

 the series really exists; and" to suggest that the series is more likely to 

 have syenite between granite and nephelite syenite than to run from 

 syenite to granite to alkaline types. This is the series shown by Pirsson 25 

 in a generalized diagram, and Bowen himself speaks of the occurrence 

 of alkaline granites and syenites between granites and nephelite syen- 

 ites. 26 His tabulated series, however, does not specify the subalkaline 

 syenites which must be the only possible syenites that could grade through 

 granite to alkaline rocks. 



The diagrams show several good examples of a series from gabbro to 

 essexite to nephelite syenite. These show that Bosenbusch's several series 

 were not really distinct. It is also rather different from any series sug- 

 gested as likely by Pirsson, on the assumption of a primary magma of 

 "intermediate" composition, such as monzonite, or the "average igneous 

 rock/' The sequence is mentioned by Bowen as a third possible se- 

 quence, 27 but even in this case he postulates a development of nephelite 

 only by the "activity'' of water in developing quartz rod's. 



The idea that nephelite develops only by an equilibrium which allows 

 the settling or separation of quartz is certainly not favored by the dia- 

 grams. Almost one-fourth of the series show a transition from olivine 

 rocks to nephelite rocks without any development of quartz rocks. Still 

 others show so little quartz that it can not account for the nephelite 

 found. This objection, to be sure, is not a rigid one, for it may be that 

 the magma was differentiated and quartz separated before intrusion into 

 the present chamber. The prominent examples, however, indicate some 

 process of development of nephelite besides the separation of quartz. 



A final suggestion from the diagrams is based on the evident tendency 

 of several series to scattering at each end. The scattering at the gabbro 

 end of the "normal" series is attributed to the accumulation of crystals 

 from a gabbro magma. The scattering at the acid and alkaline ends has 

 not been emphasized, but seems to be rather a conspicuous feature. Prob- 

 ably the best sequence for the differentiation of "primary basaltic" magma 

 into both granite and nephelite syenite would be a series leading first 

 toward syenite; from a syenitic magma the separation of granite might 

 leave nephelite syenite still liquid. Such a series would not properly be 

 ■through granite to nephelite syenite, but through syenite to both nephe- 

 lite syenite and granite. And it is to be emphasized that data for such 

 a series are not yet available, while there are some 50 or more other well- 



25 L. V. Pirsson : Rocks and rock minerals, p. 144. 



26 Op. cit.. p. ."7. 



27 Op. cit., pp. 57 and 78. 



